They probably just said "Well, it was _only_ a retrofit, and you just changed a couple of minor things for the test. Are you sure we need to spend $17 million on another throwaway test?" And the engineer said he's not 100% sure it's necessary, but it would be advisable. So the higher up, not understanding how brittle and complex systems are or how development works, decided that meant it wasn't necessary.
I think the main issue might be that there is no way to do a good test without spending millions of dollars. Which I think might be because of structural issues in the defense industry.